Allegations of a Bribe-Driven Facebook-OnlyFans Conspiracy Unsurprisingly Fall Apart in Court–Dangaard v. Instagram

The plaintiffs’ allegations were sizzling. In my previous post, I summarized:

This lawsuit involves troubling allegations that Facebook executives (allegedly, Nick Clegg, Nicola Mendelsohn, and Cristian Perrella) took bribes from OnlyFans-related entities to spike Facebook and Instagram posts that promoted competitors of OnlyFans. Allegedly, the spiking included naming the plaintiffs on the services’ lists of “dangerous individuals or organizations,” which then fed into GIFCT to block the plaintiffs on other sites too.

To support these wild-sounding conspiracy claims, the plaintiffs pointed to a BBC article that relied on anonymous whistleblowers and various anonymous tips.

This case got assigned to Judge Alsup’s courtroom. I previously summarized Judge Alsup’s modus operandi about motions to dismiss:

Judge Alsup gives the benefit of the doubt to plaintiffs on motions to dismiss, only to hammer them on summary judgment if their evidence doesn’t hold up to scrutiny.

In this case, Judge Alsup gave the plaintiffs deference at the motion to dismiss stage, and even made a dubious interpretation of Section 230 to benefit them. Want to guess what happened next?

Judge Alsup doesn’t mince words in summarizing the outcome:

This action began with spectacular allegations that OnlyFans had bribed Meta to blacklist adult entertainers who also used other competing sites. This was, in a way, to “monopolize” the adult entertainment market. After hearing this allegation at least twice, the Court instructed plaintiffs’ counsel to go present proof of such a bribe and to specifically subpoena the banks that were allegedly involved in laundering the bribe. Plaintiffs’ counsel were given the opportunity and eventually reported that they could find no proof of the bribe and withdrew the allegation. The complaint then shifted to claiming that Meta, for its own reasons, had discriminated against plaintiffs by blacklisting them. Again, the judge gave ample opportunities to plaintiffs’ counsel to prove up this claim. Again, plaintiffs’ counsel failed to find any poof. This is the basic reason that summary judgment, at long last, must be GRANTED to Meta defendants.

This is quintessential Judge Alsup. He eventually dismissed the case, but only after both sides incurred substantial expenses to address the wild claims. Judge Alsup isn’t super-generous with sanctions for plaintiffs who got his initial benefit of the doubt, and given his frustration with Facebook’s record-keeping (discussed in a moment), I can’t imagine he’ll grant Facebook any sanctions here. The result is that Judge Alsup’s “careful” approach at the motion to dismiss stage enriched the lawyers and disadvantaged everyone else.

Still, don’t overlook the key takeaways. Despite the plaintiffs’ initial allegations, there never was a Facebook-OnlyFans conspiracy; there never were bribes to Facebook executives to favor OnlyFans over its rivals; there was never a nefarious plot to illegitimately spike the GIFCT database with non-terrorist content; no one committed any crimes here. It was all unsupported claims predicated on shady anonymous origins. Nevertheless, the initial claims grabbed headlines; the denouement will make barely a ripple.

In the end, this case devolved into yet another content moderation/downgrading case, and like the others, it failed.

This is a win for Facebook, but Judge Alsup wasn’t happy with how Facebook kept records about who they placed on its internal terrorist list. He says:

due to the manner in which Meta defendants maintained the DOI List, neither side can confirm or deny whether an individual was on the List during the relevant period between 2018 and 2019….It is disturbing that Meta failed to archive a daily copy of the DOI List so that in future litigation (or consultations with federal or state law enforcement) it could identify when someone was on the list. It would have been easy to save the list at the end of each day.

You can essentially hear Judge Alsup’s teeth gnashing in the opinion. Facebook’s lack of archival records helped the plaintiffs’ case fall apart. They never had smoking gun evidence in their possession, and they couldn’t confirm their blocklisting via discovery because Facebook didn’t keep the data. I understand Judge Alsup’s point that Facebook could have and maybe should have kept better records. Then again, the plaintiffs have to prove their case. Their claims were based on anonymous claims, innuendo, and assumptions, so they failed their burden of proof.

The opinion also provides a little detail about how Facebook now places items on the terrorist block list, which gets transferred over to GIFCT. Maybe this detail has been explained elsewhere in public records, but I don’t recall seeing it. The court explains:

there is a multistage nomination process to  be placed on the DOI List, which is primarily done by a team of experts within Meta who work under Mr. James. This team then compiles a package of evidence which is reviewed by a cross-functional team at Meta which includes the DOI Policy Team, DOI Process Team, Core Policy Team, Strategic Response Policy, Legal, Communications, and several other departments. The designations are ultimately approved by the DOI director, the strategic response director, and a core policy director. If the designation decision is not unanimous, it is escalated to a more senior member of content policy leadership, who ultimately makes the final decision

That’s some serious (and expensive) bureaucracy, and yet all this procedure doesn’t fix some structural problems with GIFCT, e.g., it doesn’t ensure transparency into GIFCT’s database or the ability to identify and correct false positives in that database.

Case Citation: Dangaard v. Instagram, LLC, 2024 U.S. Dist. LEXIS 171462 (N.D. Cal. Sept. 23, 2024)